摘要
随着现代计算机图形图像技术的发展,目前的抗锯齿技术存在帧频率低、消耗内存大、对硬件要求高、算法不稳定、图像失真等问题.针对以上问题,提出基于GPU的实时抗锯齿算法,是根据几何形态学图像分析方法来区分像素边界,然后计算像素颜色的混合权重,最后对模型进行渲染,不仅可以用于前处理,而且可以用于后处理,完全独立于渲染管线,可以有效地解决多物体复杂三维场景下的锯齿失真问题,而且不会对系统造成太大的负担,是一种既快速又健壮的抗锯齿算法。
Along with the development of modem computer graphics technology, the existing anti-aliasing technology suffers with many problems, such as low frame frequency, consuming large memory, high requirements for hardware, unstable algorithm and distortion of the image. In order to solve these problems, proposes a new real-time anti-aliasing technology based on GPU. It distinguishes pixel boundary depending on the geometric morphology image analysis, then calculates blending weight, finally renders the model. It is not only suitable for pre- processing but post-processing, and completely independent of rendering pipeline. It can effectively solve the distortion under the complex 3D scene model and cost less resources of the system. It is a fast and robust anti-aliasing algorithm.
出处
《现代计算机》
2016年第2期54-57,共4页
Modern Computer
关键词
抗锯齿
GPU
几何形态学
并行处理
三维场景
Anti-Aliasing
GPU
Geometric Morphology
Parallel Processing
3D Scene Model
